THEARUPJOURNAL

JUNE 1974
 THEARUP Vol. 9 No. 2 June 1974
Published by
Ove Arup Partnership
13 Fitzroy Street. London. W1 P 680
Contents

The Open University. 2

JOURNAL Editor : Peter Hoggett

Art Editor : Desmond Wyeth FSIA
Editorial Assistant : David Brown
by F. Coffin

Paradise Circus.
by E. Irwin
8

Design of continuous 21
composite beams
for buildings.
by J . Morrison

Front cover: Symbol of the Open University ( Reproduced by courtesy of the Open University)
Back cover: Paradise Circus reference library - The circular cast iron staircase from the old library building erected in the local studies
library. Four coffers were made solid where the staircase penetrates the floor. (Photo : Logan Photography)

The Open University
Frank Cotti n
The beginning
The idea was first suggested by Harold Wilson
in a speech in Glasgow in 1963. When he
became Prime Minister in 1964. he asked
Jennie Lee. now Baroness Lee of Asheridge.
to become the first Minister of the Arts. He
also asked her to investigate that idea for
making higher education more widely avail -
able by use of technological developments.
This was for an open university teaching
through television and radio. She accepted the
challenge. and initiated an examination of the
potentialities of the scheme which resulted in
a White Paper in 1966. The University of the
Air. the title by which the project became
known at the beginning.
A planning committee set up under Sir Peter
Venables. now Pro-Chancellor and Chairman
of Council of the University, studied the early
suggestions. Their report. in early 1969. pro-
posed that broadcasting would be only one of
the media by which the university would
perform its teaching role. This report gave
birth to the Open University as we know it
today.
In March 1969. Ove Arup & Partners received
a letter from the Secretary, then Secretary
Designate. A. Christodoulou. informing us
that Fry. Drew & Partners had been commis-
sioned to act as site development architects.
and offering us the appointment as consulting
engineers for the structural design of the new
buildings on the chosen site. It is typical of the
speed with which things were happening in
those early days that we had already started
digging trial holes and concerning ourselves
with minor repairs and extensions to an exist-
ing building so that it could be made ready as
the main reception area and offices for the
Vice-Chancellor. the Secretary and the van-
guard of the people necessary to develop and
manage this new project.
Fig. 1
Architects' Strategic Development Plan

I-MU

·--
l'lft-
l-

\lftUJ---
llOW.-·
1
I 111111-

-~-
1S -"'4.laMATO
l!WI_IH_
IINTl-
l!Ullf,t(-

'J. """""'["M, ..4..0-"'te

5
•
V

·l
u-,,.._~_ _ _J•

~ --·--
__
- ---.....
-........ '--mac
~ ....-.

~-- -1111

..
I I
l
/

2 ... -· ··-
Fig. 2
The Open University site
(Photo : the Open University)

Fig. 3
Walton Hall (Photo: the Open University)

The site
The site chosen was in open countryside. four
miles north · east of Bletchley in Buckingham ·
shire. 50 miles north of London. and in the area
designated for the new city of Milton Keynes.
Readily accessible from the exit from the M 1
just south of the Newport Pagnell service
station. it is in an area of gently sloping open
fields interspersed with the occasional clump
of trees. bushes and hedges . The east boun·
dary is a minor road leading southwards from
the original village of Milton Keynes through
the small village of Walton . To the west is the
small meandering River Ousel. little more
than a stream. running through a low marshy
area in normal conditions. but liable to flood
over part of the western area of the site under
adverse conditions.
In the centre of this area. on the crest of a rise a
few feet above the surrounding area. was the
original building. Walton Hall. Not large by any
stately home standards. it was nevertheless a
pleasant and quite imposing building set
amongst some magnificent trees. forming a
reminder of the pattern of life in this area in
earlier days. The only other building nearby is
St Michael's Church. set amongst trees on the
south·west corner at the end of Rectory Lane.
which leads from the Rectory on the main
eastern approach road and forms the southern
boundary to the university site.
On one of our early site visits we noticed some
familiar temporary huts and tripod rigs about.
and wondered whether Geotechnics were even
quicker off the mark than usual in anticipating
our needs. Further investigation disclosed that
Wimpeys had been commissioned by the
Milton Keynes Development Corporation to
investigate the subsoil over the line of a pro · 3
 posed deep main trunk sewer to run in a tunnel
under the site. By a little hard talking we were
able to persuade the Corporation and Wimpeys
that a few more boreholes and samples were
neither here nor there. and got ourselves a
cheap site investigation. We found that gener-
ally over the site we had a fair depth of topsoil.
up to 0.6 m in places. overlying 2-3 m of soft
boulder clay, and then the Oxford clay. which
was stiffer but still softish in the upper layers.
These ground conditions produced no serious
problems in the foundation designs for the
future buildings. Normal spread foundation
were found to be appropriate for all the early
buildings. but the depth of the topsoil. and of
the soft boulder clay, meant that those founda-
tions tended to be costly because of the depth
of dig and the lower ground pressures to
which they were designed. Piling is being
considered for one of the buildings now under
design because of the heavier loadings in-
volved. but the site cannot be considered as
difficult in any way from aspects of foundation
design.
The first buildings
Two thing~ became apparent to us when we
first started working with the architects. The
first was that. having decided to embark on
such an unusual adventure. everyone from

Fig_ 4
The Arts Faculty (Photo: the Open University)

Fig. 5
The Operations building
(Photo: the Open University)

4
Jennie Lee downwards was desperately
anxious to get it off the ground. or on the
ground so far as the buildings were concerned.
as quickly as possible. The second was that,
since no-one else had started an open university
before. there was no previous experience to
draw on and no-one really knew precisely
what facilities were required for such a venture.
It was therefore decided. just before we
arrived on the scene. to split the early buildings
into two phases. Phase 1 was to consist of the
simplest possible buildings. to be constructed
of load-bearing brick and proprietary precast
concrete floors. based on a straight central
corridor plan and relatively shallow offices
either side. Together with the first part of the
refectory, the boiler house and a small building
to act as science preparation laboratories,
these were to be built on a design and con-
struct basis to the requirements of Fry. Drew &
Partners. by the appointed contractors. Y. J.
Lovell Ltd .. in the shortest possible time. The
second and larger phase was to include the
many other facilities which had to be worked
out as the design progressed. and which were
essential to the early functioning of the new
university.
The main Phase 1 buildings. later to be
extended to a similar plan and elevational form
but an amended structural design prepared by
us. became the faculty buildings for arts. social
sciences. educational studies. and for the
Institute of Educational Technology. In one
important respect they set the pattern for the
remaining buildings which followed. They are
basically low buildings. three storeys maxi-
mum. with an elevational treatment relying
heavily on brickwork. The Phase 1 buildings
were started with a rather harsh red local brick.
but on the later buildings a warmer reddish-
brown facing brick was used which. in my
view. blends well into the environment into
which the University is set.
Phase 2
Whilst the first buildings on the site were
simple in form. and based on the requirement
for relatively straightforward office accom-
modation. it was in the development of the
next and larger phase that the real early needs
of this special form of university had to be
determined and met.
Inevitably there was a need for more office
accommodation. but with more flexibility than
the rather rigid structural form the early faculty
buildings allowed. It was also apparent that
one of the urgent and vital needs of the new
University would be the means to process and
handle information flowing in and out on
paper. It was also evident that there would be
lots of it. Finally. there was a need for a place
of assembly. The University would not have the
need for a number of lecture rooms. for obvious
reasons. but it was important to them that they
had some accommodation on site which
would enable the academic staff and adminis-
trative staff to get together for discussion. An
assembly hall large enough to accommodate
approximately 350 was included in the second
phase. but designed as a multi-purpose space
to provide recreational facilities for the univer-
sity staff also.
The Phase 2 buildings occupied a different
part of the site. on the east side of Walton Hall.
The visual connection between the two is
therefore broken by the previous buildings on
the site and some magnificent trees which are
being preserved with loving care around the
hall itself. This meant. for the designers. that
within the overall framework of low rise build-
ings with a basically brick elevation. they
could consider the second phase without
being restricted by the rather specialized
details of the first faculty buildings.
The administrative and operations buildings
were therefore developed as 1 2.2 m wide
blocks. two and three storeys high. interrupted
by separate staircase units set at appropriate
points along their length. To give maximum
flexibility in their internal planning. the struc-
tures of the individual blocks are 250 mm deep
flat slabs on columns at 6.1 m centres in both
directions. The top surfaces of the slabs are
floated off to receive a carpet finish and avoid
the cost of screeds. The external walls run
Fig. 6
Arts Faculty block
(Photo: the Open University)
proud of the outer row of columns and are
formed of continuous strip windows with
cavity brick below. The external skin of brick-
work is carried on a continuous steel angle
over the windows. These angles were galva-
nized and painted with pitch epoxy paint
strictly following Turlogh O'Brien's recom-
mendations at that time and turning a deaf ear
to the rumblings in other sections of the design
team about the cost of these most essential
precautions.
These buildings are now basically planned and
used on a central corridor theme. with the
main services running horizontally and ex-
posed below the slab. The cast-iron down
pipes run vertically by centre columns. also
exposed. The runs of services were well
planned. and despite the lack of false ceilings
and enclosing duct-work. the interior gives an
impression of order and good design out of
austerity. The central corridors now form a
main communication link between various
different parts of the University. but in some of
the areas off this communication route. advan-
tage is taken of the flexible planning arrange-
ment by enlarging the corridor out into open
spaces to meet the needs of the particular
users.
An essential requirement for the University is
the facility to process and handle information
pouring backwards and forwards from the
students. Inevitably therefore one of the
requirements within Phase 2 was a building to
house the computer. It was found possible to
incorporate this within a wider block. but of
similar form to the remainder of the adminis-
tration buildings. But information transmission
and processing also means masses of paper.
This means that the Open University is to
become one of the largest printing organiza-
tions within the United Kingdom. and needs a 5
 Fig. 7
The Catering b lock
(Photo: the Open University)
correspondence service and printing depart-
ment of the appropriate size.
Phase 2 therefore inc luded a series of single-
storey bu ild ings alongs ide one of the main
operations b locks to hand le incom ing and out-
go ing paper and house the necessary printing
equipment. It is these buildings in particular,
and the roads and service yard required to feed
them, which begin to illustrate the difference
between the Open University and the more
conventional university scene. The architects
were left with the problem of incorporating
what are basically industrial buildings within
an academic setting. In arriving at a solution.
they rejected the standard north light form of
roof. and the adopted solution was a series of
six 12.2 m bays of single-storey steel framed
construction. with brickwork up to gutter level.
and equal sloping north and south light glazing
between each bay. The resu ltant bui ldings
blend in we ll wit h the rema inder of the Univer-
sity site, but inevitably some problems are
experienced in the middle of summer by the
occupants.
The next phase
By the time Phase 2 was nearing completion,
plans were well afoot for extending the build-
ings already built. or nearly so. Another faculty
block was required on the western side of the
site, to extend the original Phase 1 buildings.
This was built to the same basic design as
before, but with a slightly amended structure.
Also the existing refectory was extended to
cope with the larger numbers now invo lved in
running the University.
Like all good computers. the Open University
model, an ICL 1902A, started demanding
more data to feed it. At the same time the
mathematicians decided that their new build-
ing must have more affinity with the comput-
ing ce ntre than with the arts and socia l
sciences people on the other side of Walton
Hall. Two more blocks of sim il ar construction
to the Phase 2 offices were therefore designed
6 and built for data processing and mathematics.
The first of these was a near replica of what had
gone before, and was tied up to the com-
municating corridor network and the computer
building itself by one of the standard staircase
towers slightly amended . Mathematics stands
on its own. however. with an internal staircase.
which allowed the architect to introduce a
variation in the theme at one end at least.
The most important new project in this phase
of the work is the library, or at least the first
stage of the library . It is perhaps significant.
perhaps not that a site was found for it
between the arts facu lties and the rest. but
nearer and connected to the former. It was
intended that the library shou ld be built basic-
ally in two more or less equal halves separated
by an expansion joint. The first half is nearing
completion and handover now. It is a simple
building, two storeys, approximately 27.4 m by
40m with columns at 6.7m centres and a
300 mm solid flat slab. The elevational treat-
ment at ground and first floor follows the same
pattern again as the Phase 2 buildings, with
cill height brickwork and continuous strip
g lazing. The edge detail at roof level. however.
introduces a new variant in the design in the
use of a deep fair-faced concrete fascia beam .
It was origina lly intended that this should be
precast. but Y. J. Love ll Ltd. elected to cast it
in situ. and have produced a good standard of
finish throughout its length .
The temporary end of the first part is probably
the least sa tisfa ctory elevation of the building,
but if the library follows the previous pattern of
development as the rest of the site. it will not
remain visible for too long. On the same basis.
it is also unlikely that the second stage of the
building will be the same as was envisaged in
the original design, such is the rate of change
of requirem ents of the University as it co n-
tinues to develop.
One other problem started to arise on site
during Phase 2. and during this phase.
Whereas the Open University is unlikely to
experience the same problems of students
driving their ca rs around the University plain,
as occurs elsewhere, they do have other traffic
problems, which include the arrival and
departure of heavy lorries carrying tons of
paper and other supplies. Ove Arup & Partners
were therefore asked to extend their brief to
include the design of the service area. roads
and ca r parks.
Part of this work had been started to designs
prepared by others. Our first task therefore was
to link up these to a new road round the
western side of the site to produce a circular
ring roa d. Two problems arose here. Because
of the proximity of the western part of the
River Ousel. the route took it inevitably
through the area liable to flooding. At the same
time, and arising partly out of its location, this
was an area with a fair depth of topsoil over-
lying a very soft clay. Frank Fowler and Ray
Ham have been concerned w ith the design and
supervision of the construct ion of this road and
it is performing we ll. The Un ivers ity are aware
that at its lowest point it is li ab le to fl ood ing.
but have accepted this as a calculated risk.
Some trouble was experienced with the
service yard adjacent to the single-storey
co rrespondence and printing department. It
was apparent that the traffic in this area was
both heavier and greater than had been antici-
pated originally. Again Frank Fowler produced
a design for the reconstruction of this area.
which is one of the focal points for heavy
traffic on the site. Th is has been completed to
form a first class service yard in this area.
The immediate future
At present the fac ili t ies on site tend to be
weighted towards the facu lt ies of arts, soc ial
sciences. educationa l studies and mathe-
matics. We are at present engaged on the
design of the first stage of the science and
technology building, to be located on the
north-east corner, beyond the computer area
and mathematics building. Not only will this
building house the staff of this faculty, but it
will also provide a research laboratory for
those staff and for their research students.
The additional servicing requirements arising
from laboratory use mean that the structure of
this building will partly break away from the
pattern developed on Phase 2 with a closer
spacing of co lumns along the length of the
blocks to allow greater penetrability of the
horizontal structure. This will undoubtedly be
reflected in the external treatment. but the
basic principles of low-rise construction with
a basically brick treatment will be followed.
The first stage of the science and technology
building is planned to be followed by two
further stages. which in terms of the present
planning will make it one of the largest of the
faculty buildings on site .
Fig. 8
Administration building
(Photo: the Open University)
Although it was initially conceived as 'the
university of the air'. radio and television in
fact occupy only a small amount of the
student's time. and the most important element
in the system is a correspondence package
which most of the students receive weekly.
The television and radio programmes forming
part of the course are recorded at the BBC
studios at Alexandra Palace. The lease for
these studios runs out in 1977. and the Univer-
sity has to find other accommodation in which
to produce its programmes . It was decided
therefore to build new studios, with the co -
operation of the BBC. on the University site.
The architects appointed for these stud ios are
the Ware MacGregor Partnership, who have
considerable experience in television and
broadcasting work. The preliminary designs
have been prepared. and it is hoped to start
construction in March 1975. The scheme in -
cludes an extension of the Phase 2 office
blocks in similar construction, to provide the
inevitable requirement for more office accom -
modation. This leads into a large one - and two-
storey studio complex more than 200 feet
square on plan, housing the two studios and
Fig. 9
Catering block (Photo : the Open University)
an additional small drama and music studio.
the control rooms. rehearsal rooms. property
and scenery stores. and all the various confer-
ence and technical accommodation required
to process and package sound and television
programmes so that they can be delivered to
transmitting stations and then on to the
students over the air.
The technical requirements for the individual
parts of the studios are so specific that they
tend to dominate the architectural and struc-
tural design. The height of the buildings and
their elevational treatment will be compatible
in form with the existing buildings on the site.
By its very size. however. the studio complex
will undoubtedly form a dominant feature of
the site.
As structural engineers for the studios. our aim
from the beginning of the design process was
to produce a structure flexible in use. but keep-
ing as much out of the way of the inevitable
services. air handling ducts and electrical and
electronic hardware as possible. The required
column spacing produces a more rectangular
grid. To maintain a flush soffit. give easy fixings
for cables and allow the maximum penetra-
bility of the floors. we are developing further
on the ideas which Nigel Thompson started on
York Hospital of a flat slab with Lignacite
blocks inserted where the heavier concrete is
not working hard. This form of structure will be
used over the technical areas wherever it is
appropriate. Over the property and scenery
stores the roof will be in lightweight construc-
tion supported by steel. The studios themselves
require special treatment for acoustic reasons.
and here a double skin construction is being
considered with S1porex autoclaved light-
weight concrete supported by structural steel-
work forming the roof over.
The future
One of the exciting things about working on
the Open University site is that the present is
always quite different from the earlier fore-
casts. When we first started it appeared to us
that we were building something not very
different from a quiet university campus
around Walton Hall. As the Phase 2 buildings.
and subsequent extensions. came into the
picture. the scene appeared to change subtly,
as the size of the operations required to process
and distribute the packages of learning
material to the students became apparent.
Undoubtedly, the completion of the studio
complex will change the scene again. But
already the need is becoming apparent for a
large warehouse on the site. which may well
be under way whilst the studio is being built,
so the future pattern of the whole site con-
tinues to change to the extent that it remains as
uncertain as before.
I should perhaps emphasize finally that this
article has been written solely from the point of
view of one involved only in the buildings at
Milton Keynes. from the beginning. Through
its widespread regional offices and study
cent res. the total University is much wider than
those buildings suggest. Based solely on
student numbers. with 38.000 at present, the
Open University is now the largest university
in the country. This is obviously an invidious
comparison. because its research and other
activities cannot begin to co mpare with the
others. But when one considers that the first
members of staff were appointed in January
1969. the royal charter was presented in July
1969 and undergraduate teaching started in
January 1971. then the progress to date of the
Open University can be seen in rather startling
perspective.
7
 Paradise Circus
Ernie Irwin
Introduction
The Paradise Circus complex forms the heart
of Birmingham's Civic Centre. It is situated
between Birmingham 's Town Hall and Art
Gallery on one side and the City Council's
Administration Centre. Baskerville House. and
the new Repertory Theatre (Job No. 1976) on
the other side. The complex consists of a
major new reference and lending library, a
Later phases
school of music, lecture theatres. offices.
shops, public houses and public areas. These
are arranged over a bus interchange and
underground car parking. The bus interchange
in turn is built over the city's new Oueensway
Ring Road which passes through the site in an
underpass (Figs. 1-3) . All of the se buildings
are being constructed within Phase A of the
project and the last of this stage will be com-
pleted when the old Birmingham Library on
the site has been demolished . The new
libraries and school of music have now been
opened (Figs. 4-6).
Design of the project commenced in 1964 ;
preliminary construction started in 1968, with
the main contract for the buildings commenc-
ing in 1969. Detail design continued until 1971
and in all some 3,000 structural drawings were
produced for the project. It is an extremely
diverse project to describe and this article is
necessarily lengthy. Consequently, the paper
has been divided into sections as listed below
for easy reference for those with interest in
special aspects of the work.

(1) Introduction Fig.1 Oueenswav

(2) Structural materials Plan at basement level Road tunnel
(3) Foundations
(4) Underpass and gyratory road
(5) Concourse
(6) Reference library block
(7) Lend ing library block
(8) Link blocks
(9) School of music
(10) Contract arrangements
(11) Setting out
(12) Acknowledgements.

The north west end of the site. known as

Phase B, will consist of further underground
car parks and other civic buildings, but no
programme is yet established for that stage.
Planning for comprehensive development of
the site for civic use commenced in 1964,
although the general traffic plan to run the
Baskerv ille
A38 through the underpass and form a gyra- House
tory road around the site was already deter-
mined .
The main library was located in order to avoid
New
it sitting directly over the underpass, although C 1v1c Centre
many of the ancillary buildings are necessarily Area
on the underpass route and there are in
general three levels of different fun ct ions
across the main part of the site, i.e. bus inter-
change, concourse and library.
The various buildings have different uses and
accommodation needs and according ly a
particular constru ction system was chosen
for each separate building type. Different
structural systems were economical ly viable
due to the large size of each individual build-
ing. but in the service areas the structural
grids were determined both by the building
above that area and by the bus interchange
layout.
The total building works were valued in excess
of £Sm. in 1968 and the Phase A building Fig. 2
project was let at a value of £5m. and the Plan at concourse level
8 roadworks at £1.5m. in that year.
 Structural materials
On a project so large and diverse most struc-
tural materials were used. However. notwith-
standing the fact that several hundred tons of
structural steel were incorporated. the build-
ings are primarily reinforced in situ concrete
w ith precast concrete cladding. Almost every-
where within the buildings the concrete is fair
faced and exposed to view.
All in situ concrete for the building is
30 N/mm 2 compressive strength at 28 days,
and was made from a Trent Valley 20 mm
natural aggregate. The concrete was site
mixed to ensure consistency of strength and
colour. as the colour was critical due to
columns and most walls being grit-blasted. In
addition stocks of aggregate were stored at the
pit to allow supplies to continue when varia-
Baskervil le tions occurred in the material.
H ouse
Four types of concrete finish were employed:
(1) Plain concrete from plywood forms used
in service areas. below ground and on
hidden surfaces
New
/1 Ci vic Centre (2) Grit-blasted concrete to approximately
\J area 3 mm depth on most vertical surfaces
throughout the libraries and link blocks
(3) Grit-blasted ribbed concrete on service
core walls
(4) Smooth fair faced concrete from glass
reinforced coffer moulds on reference
library floors.
Very consistent concrete was obtained in
terms of strength and colour. but dark blem-
ishes occurred on a number of grit-b lasted
columns due to knot ho le repa irs in the ply-
wood . or the vibrator touching the formwo rk.

Fig. 4
View of libraries looking west from the
archway over Edmund Street. The main
Fig. 3 entrance is through the lending library in
Roof plan the foreground

9
 .J
I I I I I I II I I I I I II

In all cases joints in the plywood show. and a Fig . 5

better arrangement of plywood joints would View of reference library with Baskerville House and Hall of Memory in foreground
have improved the appearance in many cases .
Precast cladding and roof tiles were manu -
factured using a Hopton Wood limestone
mixed with Derbyshire Spar with white cement
and these were lightly grit-blasted so as to
simulate the Portland Stone used on the exist -
ing civic buildings.
Reinforcement was generally high tensile
steel with mild steel links. although building
bye laws limited stresses in foundation re -
inforcement to tho se of mild steel. Structural
steelwork included plate girders approxi-
mately 2 m deep.

Foundations
The site is founded on Birmingham 'rock sand '.
whi ch is a non - cemented sandstone capable
nevertheless of high bearing pressure. The
ro ck sand had been proven by the Local
Authority's borehole s taken for the Inner Ring
Road.
It was decided to found all the buildings on pad
footings at a bearing pressure of 750 kN/m 2.
However, preliminary excavation of the under-
pass below the site revealed the presen ce of
thin in clined strata of marl rising from north
to south across the site. In some cases these
layers occurred immediately below foundation
level. As a result the foundations were lowered
where feasib le and in other cases the founda -·
tion pressure was redu ced to 430kN/m2.
When the marl layers had been discovered
further site investigation was quickly under -
taken with one of our engineers continually Fig. 6
present. and by careful observation it was School of music with shopping centre beneath . Note recital room adjacent to busy road
possible to detect the layers of marl which with underpass beneath
10 would and did escape a normal investigation .
 ,, Precast rafters and

q~ ' \ ~p roof tiles

I I I 1 Coffer slabs
•
I I ..------------. - I
••
Precast cladding

• I
I
Reference

library
I
I I-...__ I rl
""' u
'I' ',, II .
Lecture Le~ding 1:,brary
t heatre
Gyratory
road

Phase B
II
Service I
----- Pla nt
I
Concourse
11

Bus interchai
Gyratory
II
II
I
II
u
Children's library
Car
I
....,._
road rooms I road
park .Qilil,,,

o II I .
a Unde. -pass -i. .I • II - -
I ,I Pedestrian tunnel I I I I I I I
I I Ill
{_ r.1.-..1., Services tunn el rJ.. .1 , .-..1...1., r ..J.. _ ..1.,
'+
Fig. 7
Paradise Circus - Phase A. Structura l section

From the information found the strata were

E
E plotted across the site and all further founda-
tions were adJusted accordingly.
0.4
Brass settlement plugs were placed in some of
the columns at bus interchange and conco urse
level as the project commenced and accurate
0.3 Levels taken at settl ement measurements have been observed
c t hroughout construction.
positio n 8
" The average settlement on completi on is
~;; 0.2 pos111on 3
approximately 4 mm and a load settlement
V)
4 position 9 graph is shown on Fig. 8.
3
Rock sand. though uncemented. was found to
0 .1 remain stable in deep excavations without
shoring. even when exposed for several
months. However. we insisted on protect ion of
the banks of the 14m deep excavation around
60 80 90 100 the existing library wh ich was to rema in open
for several years. Grout-covered hessian was
Fig. 8 Percentage design load
used to prevent erosion by run-off of surface
Reference li bra ry load/settlement graph water.
Underpass and gyratory road
Colu mn load Col umn load The new Birming ham Inner Ring Road was
Bus interc hange level pl anned through the site as an underpass
below the bus interchange level . It was
. . - - - -In situ beams under designed as can til ever retaining wal ls bui lt in
column loads (up to
..--- - - - ' - - ' - f - - - - - -- ---~~---+- - ~ 1.524 m deep) open cut with standard prestressed precast
inverted tee beams and an in situ toppi ng
forming the roof ( Figs. 9 and 10).
While the main library was sited so as to avoid
Carriageway Carriageway R.C. retaining the underpass. many of the ancillary bu ildings
R.C. wall were necessarily supported on it and in th ese
wal l locations large in si tu reinforced concrete
R.C. wall
beams were used to carry the column loads to
the main retaining wal ls of the underpass. The
Road pavement Road pavement maximum load carried is 9000 kN under one of
the lift cores wh ich supports two link bl ocks.
With experience of this construction. it wo uld
li kely have been better had the underpass been
designed as a double box culvert in reinforced
Fig. 9 concrete over the length of the site. This was
Underpass cross-section not possible as we were given authority to
make only local alterations to the standard
Public Works Department prestressed beam
design.
Provision has been made on the underpass
roof. and adjacent to it. for loads from Phase B
of the project.
The other major road affecting the site is the
new gyratory road which forms the north.
south and west boundaries. but on the east
side dips throug h the complex at bus inter-
change level and beneath the lend ing library
and concourse. Fig. 7 illustrates the re lation-
ship of underpass. gyratory road and mai n
buildings.

Fig. 10
Underpass during co nstruction
11
 Walls and columns in reinforced concrete were
constructed on either side of the gyratory road ,--,-..--
to provide support for the concourse areas and
lending library above it.
Prestressed precast concrete beams with in
situ concrete topping span over the gyratory
road below the concourse area. Where it
' r - - Column sra,rers
welded to beam

passes under the lending library, 1.2 m deep

steel beams were used in groups of three to
support the building which varies in height
from three to five storeys.
Construction depth at this point was limited to
I -,; . .. 1• :~ ..
~·
330m m
1.2 m because the levels of the gyratory road
.
Ti' 'n' 'n'
i 230mm
~

could not be adjusted without seriously affect-

ing the road layout. nor could the levels of the I I
lending library alter without affecting the total .- >-- Web stiffener plares
building development.
The problem arose of how to carry a re- E
N
inforced concrete building with core walls N
Diaphragm plates
rising to five storeys over 15.2 m span with an ~

ar 1.37m centres
available depth of 1.2 m without excessive
deflection. An additional complication was the I I
short time available for constructing this deck
as it was sandwiched between the road being
'
made available by the Public Works Depart-
Fig . 11
ment and the requirement for it to be opened
Section through triple girders spanning New Congreve Street
to traffic a number of weeks later.
The solution found was to use Autofab steel
beams in groups of three which were united by
Fig.12
steel diaphragms at 1.37 m centres and flange
Concourse area - View of reinforcement in
cover plates (Fig. 11 ).
2.67 m deep beam grid which carries one
Braithewaite & Co. Ltd. fabricated the triple of the library service cores
beams completely at their works and brought
them to site as 30 tonne units.
The last portion of the gyratory road is
presently on a temporary alignment and will be Fig. 13
completed as soon as the old library is South east corner of reference library with
demolished. Big Brum. the Council House clock tower.
in the background
The road construction and the underpass
structure were designed by the Public Works
Department but the amendments to the under-
pass structure to enab le it to carry the building
loads and the structures around the gyratory
road were designed by ourselves.
Concourse
The entire development is served by an exten-
sive multi-level public concourse at general
ground level affording access across the site as
well as to individual elements of the project.
It is intended that in addition to general use by
the public, the concourse should be used for
mounting of exhibitions including those of
sculpture. A further requirement is that this
level should be capable of supporting fire
appliances.
The grid of columns supporting the concourse
is basically the 10.87 m square grid derived for
the reference library and extended outwards.
The pattern becomes irregular. however. where
the column positions are dictated by other
buildings above or by plant room and service
areas below. These parameters. together with
the incorporation of ornamental pools. called
for the adoption of substantial design live
loads. A solid 450 mm deep reinforced con-
crete flat slab was chosen as the typical struc-
ture for this area . However. over the gyratory
road the concourse is constructed of pre-
stressed bridge beams and with structural steel
as described in the section dealing with the
gyratory road.
Two of the main cores providing lifts and sta irs
through the library commence at concourse
level because of the bus route immediately
below. These cores attract a very substantial
loading through the library structure and con-
sequently major supporting beams are required
at concourse level. Due to the presence of
internal walls in the main core. a grid of beams
was necessary to collect the loads and these
beams were designed using a grid analysis
taking a torsional beam stiffness of 50 per cent
of the theoretical. Even with this factor
extremely high torsional moments were in -
duced. The resulting beam system is 2.4 m
deep and Fig. 1 2 shows some of the reinforce-
12 ment within these beams.

Fig.15
Reference library - Early printed book and fine printing area
view through to the local studies library. Air conditioning duct
within grit blasted column, right foreground
The en tire concourse when completed will be
some 230 m long and as most of it is exposed
directly to the elements, considerable thermal
movement was an ti cipa ted. Due to the com-
plexity of the co ncourse area w ith pools, steps
and emergency access roads, it was difficult to
introduce movement joints in a regular pattern.
In general, movement joints have been located
at approximately 40 m centres in each direc-
tion while the main library, which 1s 60m
square at concourse level, has been isolated by
a movement Joint around its entire perimeter.
Reference library block
The structure of the library block was deter-
mined by many factors, but the maior influ-
ences were the inverted pyramidal shape of the
building (Fig. 13), the central void through
each floor and heavy floor loadings, particu -
larly in book stack areas.
The library is designed throughout with a
com mon module of 1 ft. 6 in. with column
spacings of 18 ft.* 1n the lending library and
36 ft. in the reference library. Book presses
are spaced at 4 ft. 6 in. centres in the closed
access areas. The shelving and consequently
the internal panelling is at 3 ft. centres for
maximum economy of shelf design. The build-
ing elements, wall panels. windows and
coffered f loors are on a 4 ft 6 in. grid which is
a satisfactory size for the constructional design
of these elements and 1s suited to the aesthetic
scale of the building. This dimensional co-
ordination produces a strong architectural
unity between the building structure and the
interior furnishings and fittings (Figs. 14 and
15).
A basic column grid of 36 ft. in each direction
was adopted as it suited both the library
planning as well as the bus interchange area
beneath. without being excessive struc-
turally.
* This proJect was designed in Imperial units.
The common module size was 1 ft. 6 in. and
other relevant d1mens1ons given are multiples
of this.
Fig. 14
Reference library - Social science
reading area looking down through floor
void from fine arts
The position of the central void in the library
varies from floor to floor (Fig. 16), so that
cantilevers of various lengths occur on the
inner perimeter of the building in addition to
cantilevers on the outer perimeter at the upper
storeys. These cantilevers are subject to heavy
and concentrated loads, particularly in book
stack areas and at external corners, and
required a rigid floor structure with good load
spreading characteristics.
A coffered floor slab system satisfied all these
various constraints and in addition suited the
basic square grid while providing a fair faced
soffit. Consequently, a 2 ft. deep coffered slab
of 4 ft. 6 in. module was chosen for all the
main library f loors (Fig. 17) .
The individual coffers, each 4 ft. 6 in. square,
are a non-standard size but. due to the repeti-
tion involved. it was an economical proposition
to have glass reinforced plastic moulds
purpose-made for the project.
Fig. 16
Reference library - Science and technology
library looking down to the history library.
The double storey height glazing faces on to
the central void
Fig.17
Reference library - View of the underside
of the coffer slab from the public concourse
area showing the Junction with one of the
main cruciform columns
13
 4 .8m

=Ix,
~10mm
toleranc e
to allow
for X, and X 2

Section
Detail
Fig.18
Reference library section showing floor deflections

Cen tre line ribs on 1.3 7 x 1 37 m module

Coffer
mould

Coffer
mould Centre line bolts
Bea ms
can be
removed suppo rting '- 150 x 75 mm bearers
leaving coffer Cheeks
prop supports moulds bolted
Beam to Beam -type Acrow prop
supporting bearers
co ffer
As used
moulds
Fig.19 Arup proposa l
Reference library coffer mould supports
Sec ti on 1- 1

Fig. 20
Reference library - Entrance to science and technology library on the 5th floor with marble bu st of Samuel Timmins.
Note junction of ribbed grit bla sted wal l and smooth co ffer cei ling
14
The coffer floors. while giving an impression of
uniformity, embrace structural arrangements
that are extremely varied as no two floors are 1.37 rn
alike. due to the varying cantilevers on both
inner and outer perimeters. The floors were
analyzed for bending shear and deflections on
a computer. using a grid analysis. and the unit
area taken was the half width of the building
by two column bays long. To incorporate this
area each pair of ribs was taken as a single grid
line and further analyses on a smaller grid were Sectional pla n
carried out around the column heads where
shear stresses are high.
The coffer voids are used to house the electric
lights and the electric service feeds down
through the coffer topping.
The cantilevers around the outer perimeter
reach a distance of 5 m at floor six which. on
the four corners of the building, is a cantilever
of 6.7 m. As a consequence substantial
cambers were needed on these cantilevers and
Dry pack
at the corners this reached 50 mm.
Stainless steel
In addition. conditions of live loading could
cause the floors to open or close relative to
each other and it was necessary to detail for
this movement. both in the partitioning and in
the clerestory windows ( Fig. 1 8). Detail

Due to the special construction and cost of the

moulds it was appreciated that they would be
required for re-use before the concrete could Fig. 21
be depropped. We devised a method of release. Cantilever precast panels on reference library
as in Fig. 19. in advance of the contract. In the
event the contractor used a system which
permitted the entire propping beam to be
removed by leaving a prop at the junction of
the ribs . This method, however, allowed grout
leakage at the junction of the fibreglass mould
and timber insert which we were unable to
eliminate throughout the job. and. unfortu-
nately, the very smooth coffer surfaces had to
be touched up at their junction with the recess
in the rib . The specification laid down strict
requirements for rigidity within the moulds and
a sample mould was carefully examined. The
very regular coffer ceilings which were
achieved are to a large extent due to the
quality of the mould that was thus obtained.
The recess within the rib of the coffers serves
several purposes and it is carried throughout Architectural
the library coffer floors. even in locations panel
where the coffers were made solid due to high E
E
shear or other reasons. The primary purpose of
~ Services
the recess is to house partition heads and allow
differential movements. In addition it masks
"'
the junction between coffer moulds. Around
the perimeter of the building the windows are
also housed in the rib recess (Fig. 16).
The typical superimposed load was 4.8 kN/m 2 • Top of slab
but this increases locally to 11.2 kN/m 2 in areas
of rolling book stack loading. Assembled section

Even though this relatively high loading was

required on 36 ft. spans. the storey height.
and thus the structural depth. had to be mini -
mized for planning reasons. Consequently, the
choice of a 2 ft. deep coffer slab was a
particularly happy one in that it eliminated the
need for any structural down stands within the
relatively short floor to floor height of 3.5 m
(Fig. 20).
The cantilever precast cladding to the outer
perimeter is shown in Fig. 21. Each unit was
fixed to the floor slab with two stainless steel
bolts with levelling nuts, and erection using
diagonal props went very smoothly. Addi -
tional fixings to give continuity between pre-
cast units were fitted through the lifting holes
that had been cast in the side flanges. The
inner perimeter precast cladding is shown in
Fig. 21 a.
The libraries are air-conditioned using a high
velocity induction system. The air for the
building feeds vertically from the roof through
ducts formed within the cruciform columns.
Two legs of each cruciform are hollow and it
was intended that the hollow shells should be
Fig. 21a
Reference library inner perimeter cladding units
precast. However. in order to obtain matching
concrete the contractor elected to cast the
projecting legs in situ 100 mm thick and this
was successfully achieved.
At each floor level air is fed horizontally to the
centre of the rooms by ducts passing through
the coffer ribs (Fig. 22). Although it is not
generally desirable to pass service ducts in a
structural zone. the holes formed a simple
direct route . identical at each column location.
It was carried out without difficulty apart from
the structural analysis and made a consider-
able saving in height in the building.
The plant rooms. together with archives and Fig. 22
staff areas. are located within a raked hipped Reference library - View of coffer ceil ing
roof. The roof is constructed of precast con- showing coverings to air conditioning duct
crete rafters spanning between in situ con- within coffers leading from a cruciform
crete ring beams around the outer perimeter column
and inner void . The rafters are spaced at 1.37 m
cen tres and are clad with white conc rete ti les 15
 (Fig. 23). Each rafter is Y-shaped in section
and the valley of the rafter acts as a gutter.
The white concrete tiles are 1.37 m square and
50 mm thick with reinforced ribs. They are a
development from those used on Dunelm
House (Job No. 1534).
On advice from the Cement and Concrete
Association only the perimeter ribs of the tiles
were reinforced and the 50 mm thick units were
unreinforced. They are prevented from sliding
by means of lugs projecting into the precast
rafter and their dead weight prevents uplift.
Between the concourse and the lowest refer-
ence floor are located two levels of offices and
other library facilities. These floors are gener-
ally about 10 m wide. inserted between the
main columns. and associated with them are a
Guner Precast roof 11h11 number of elevated. walkways. ramps, stairs

I and escalators. In order to achieve a clear

articulation of the various elements of the

r
design. the floors. ramps and walkways were
designed as separate structures supported on
corbels cantilevered out from the columns.
Lending library block
The lending library. although connected to the
main library. yielded a different structural
so lution due to its position over a roadway and
the fact that smaller floor spans were accept-
able. The lending library varies between two
and five storeys and is situated over the gyra-
tory road and underground car park. A struc-
Pr.call r•h•r•
tural column grid of 5.5 m x 5.5 m was accept-
able to the planning requirements. and the
floors were designed as 200 mm flat slabs
(Figs. 24 and 25). This relatively light struc -

Fig. 24
Fig. 23 Lending library - Internal view from 3rd floor
Reference library precast roof assembly near main entrance

16
ture was designed to reduce the overall load -
ing on the beams spanning the gyratory road
and basement car park. Live loads are
4.8 kN/m2 generally with 11.2 kN/m 2 in rolling
book stack areas. The children's library is at the
rear of the lending library (Fig. 26).

Link blocks
The link blocks are two- and three-storey
buildings of offices wh ich connect principal
buildings and span over roads and concourse
areas. The spans in Phase A of the project
range from 10 m to 28 m and the object of the
design was to produce a un iform structure
which could be built over various trafficked
areas. The design will also fit th~ spans and
site conditions in Phase B of the project.
The floor levels in the link buildings had to
match the levels both in the existing and new
buildings to which they connect and this
resulted in a maximum of 600 mm structural
depth between storeys within the links.
To achieve the spans required. much greater
structural depth was needed . and this was
obtained by suspending all the loads from roof
beams which could project upwards and had
no depth limitation.
Each building is carried by two longitudinal
steel plate girders. encased in concrete from
which the rest of the building is hung

Fig. 25
The readers· lounge in the lending library.
Note the grit blasted columns and ribbed
grit blasted walls

Fig. 26
Children's library - This is housed beneath
the lending library

17
 (Fig. 27). The hangers from the roof beams are
13.4m
Macalloy steel rods at 5.3 m centres at low
stress which are fire protected by precast con-
crete casings. Precast concrete cross beams Steel truss
span between the steel hangers supporting
standard precast prestressed hollow concrete
~
·-~
-__
,
./ '
~
/
~ ~
floor planks with an in situ topping. ' Plate girde r 2m deep
Precast beam with in sit u casing
The upper floor of each link is clad in white
J

~
precast concrete storey height wall panels. Precast cladding
~- ---i;,_ ,_..:.
Cumulative deflections were calculated for
each link. and the tolerances in the precast
·-~ L..r.:r I•

construction were arranged to suit the deflec- Macalloy bars

tions. L-- - with precast casing
The standard link design is for spans up to at 4.88 m centres
22 m. Several variations from the standard link
have been necessary to fit special conditions.
such as a third storey or very long spans. ~ - ~

I I
The art gallery link had to span the gyratory I

road close to the junction with Great Charles

Precast beams
Street. and consequently. the span reached at 4.88 m centres Glazing
28 m. This span. if carried in the normal way,
would have involved roof girders some 3! m Macall oy bars
deep which posed difficult fabrication and
erection problems. Fortunately. at this location
the level of the road was such that ample fjff --',fl il
headroom was available below the lowest
floor and so a second pair of main girders was
introduced to carry this floor. Normal 2 m deep Precast beams
girders at roof level were thus sufficient to
carry the top storey over this 28 m span ( Figs. Fig. 27
28 to 30). Typical link block showing structure suspended from roof girders. span up to 22 m
The continuation link from that serving the art
gallery carries a pedestrian walkway over the 13.4m
bus interchange below. which effectively
made the building three storeys. For visual
reason s it was appropriate to continue bottom Steel tru ss
girders as in the art gallery link. and these
girders were used to carry the suspended
walkways as well as the lower storey of the
link (Fig. 31 ).
The town hall link. which will be constructed Precast
after the old library is demolished. will again be c laddin g
a special case as it was decided not to make a
physical connection with the town hall but to M acalloy bars
return the link to form a U-shaped plan. The with precast casi ng
at 4.88 m centres
link structure was not conceived to serve this
plan shape and special adaptation of the
structura l system has been necessary.
The hangers encased in precast concrete
posed an interesting erection problem and the
assembly is shown in Fig. 34. Precast beams
Glazing at 4.88m cent res
The hanger rods were first hung and then the
casing was slid on and held up by a nut on the
main rod. A steel location plate was then
added before the precast floor beam was
fitted. Macalloy threaded couplings were used
to add the lower hanger. A group of spring
washers in the assembly held the casing firmly
after the rods extended under load.
At the bottom of the whole assembly a plate
washer and nut. together with locking bar. was Fig. 28
!;'late girder with
in situ casing lJ I Precast
edge beam

added. Art gallery link block 28 m span. with additional girders to support lower storey

Fig. 29 Fig. 30
The main girders to the art gallery link during erection over the The art gallery link block which spans 28 m
roadway. The stiffeners indicate the hanger positions for the
lower floors

18
 School of music 13.4 m
The school of music building is at the south
end of the site bordering on Paradise Street Steel truss
(Fig. 6). The concourse level and the base-
ment comprise a shopping centre and four
upper storeys form the school of music. The
Pl ate gird er 2 m deep
main block of the school of music is 1 2 m wide with in situ casing
and 74m long.
The lower three storeys are constructed Precast cladding
entirely in reinforced concrete consisting of a
simple two-way spanning slab on beams and
columns . Concrete surfaces in the public areas Macalloy bars
have exposed aggregate surfaces. Bush ham- with precast casing
at 4 .88 m ce ntres
mering was chosen due to the public nuisance
caused by the dust from grit blasting and t his
has worked out successful ly as the trans ition
from bush hammered areas to grit blasting on
the rest of the project is not noticeable.
The upper storeys of the school of music are
largely divided into practice cells. Precast beams
at 4.88 m centres
The need for sound insulation led to a load -
bearing brickwork structure over these three
upper storeys. Each cell is divided by a load -
bearing wa ll but is further separated by an
inner skin brick wal l, mounted on rubber pads
and with its own cei ling (Fig. 33).
The recital room is mounted on stilts and is
i Pl ate girde r wit h
Precas t tee beams
situated on top of the underpass at the Precast in sit u casing
extreme south east corner of the site with edge beam
continual traffic noise from two sides. Conse- I"-- - M acal loy bars
quently considerable mass was required for with precast casing
acoustic reasons and the building was
designed as a concrete box of 225 mm re- Precast units Precast bal ust ra de uni t
inforced concrete wal ls with precast cladding
panels. The plan area is 1 2 m by 1 8 m.
The roof of the recital room is a coffered slab
using standard moulds and the soffit is
painted. Despite the proximity to very heavy
traffic the recital room is successfully isolated Fig. 31 Precast beams
from extraneous sounds (Fig . 34). Link block 7, top storey suspended from roof girders, walkway suspended from lower girders

Fig. 32
Schoo l of music - recita l room. The structure is of 225 mm concrete wal ls with roof slab formed from standard coffer moulds

19
 Contract arrangements
The underpass through the site and the gyra-
tory road for Birmingham Public Works
Department carried a Department of the
Void Fibreglass insulation material Environment grant and thus had to be let as a
separate civils contract. On the other hand the
buildings contract under the authority of the
'. t "
Birm ingham City Architect drew ot her Govern-
ment grants and was let separately by com-
petitive tender using the special Birmingham
Timber joists supported
on block 'cel l' walls, building contract.
ceiling - plaster on A detailed phasing arrangement had been
expanded metal
worked out between the two contracts as they
had common boundaries and overlapped in
some places. As a result of various factors and
Sand / cement screed wi th
largely due to late diversion of services, the
vinyl tile finish '- phasing did not go accord ing to plan and this
greatly comp licated admin istration of the con-
......... tracts .
At the time of tender in 1968. competition was
very great and the building contract was
Expanded polystyrene
quilt
awarded to the lowest bidder. Throughout the
contract a great deal of attention has had to be
devoted to claims and possible claims which
have been a major distraction from the main
job of creating the comp lex.
Setting-out
Fig. 33
Two main contractors on site at the same time
Section through typical music practice room
is never an easy problem, but when they are
both constructing foundations for the same
building with one commencing 18 m below
ground at the side of an underpass and the
other operating at ground level. several hundred
lonom nanee ol ,oof
cil•t•ttWcief feet away, the problem of checking setting - out
becomes formidable. We decided that normal
gro und-mounted setting-out po ints wou ld
soon be uprooted by either contractor and
become unre liab le and a nove l solution was
,........ found . The site was divided into an approxi -
:sz: mately 60 m grid in two directions at right
angles and a survey firm was appointed in
advance of the contracts to set out these grids
between the roofs of the surrounding build-
ings. Ea ch contractor was then instructed to
-'+-- -+-- 35mm M.caHov t,.e1 set-out his portion of the building from these
Pt.CHtC0rte1•teCHing _
10M.cellovber accurate ly positioned high level grid lines.
This method proved very successful. and on ly
one column was found out of position.
The building works had been planned as
parallel to the nearby Baskerville House. On
the other hand the underpass, planned by the
City Engineer. was related to the national grid.
As many of the building columns were on the
roof of the underpass it was necessary to
relate the two setting-out grids. Relative bear-
ings were obtained and a computer pro-
gramme used to predict building grid refer-
ences at intervals along the three walls of the
underpass .

Acknowledgements
The structural design was initiated in 1964 by
Frank Coffin and John Martin . The Birming-
ham offi ce was set up in 1968 to co ntinue the

~
design of the proje ct and Ken Anthony and
Ernie Irwin jointly led the design teams from
-· I . London and Birmingham respectively.
~ Architect:

· · · = ··.
i
I,
t l
I

I
I
:
_ L. 1
I

I
I
:
I

:
J
I
1
i
,
lowerpr..:Ht ftoot
The John Madin Design Group in ass ociation
with Birmingham City Archite ct
Mechanical and electrical engineer:
R. W. Gregory & Partners
Quantity surveyor:
L. C. Wakeman & Partners
Resident engineers:
Matt MacKay, Bob Astley, Alan Tricklebank
Main contractor:
Sir R. McAlpine & Sons Ltd.
Contractor for underpass:
Fig. 34 C. Bryant & Son Ltd .
Link block s typi cal hanger assembly Underpass and road planning:
Birmingham City Engineer and Surveyor
Precast cladding supplier:
Photos in this article: W. Sinda ll & Sons Ltd.
Figs. 12 & 29 by Ernie Irwin. 17 & 22 by Malcolm Jordan . 4. 5. 6. 10. 14. 15. 16. 20. Structural steel sub-contractor:
20 24. 25. 26 & 32 by Logan Photography. 13 & 20 by John Whybrow Ltd. Brai thwa ite & Co. Ltd,
 which is assumed to be cracked. This has the (2) If bd,< aA,< (bd,+ 2aA,) then the plastic
Design of continuous advantage of reducing the bending moments neutra l axis is within the top flange of the steel
in the columns and can easily be achieved by beams and
composite beams substituting the equivalent stiffness of the M, = Y,[A,dc-brdn(dn-d,) ]
composite beam in the framework analysis.
for buildings The equivalent stiffness can be expressed as (3) If a (A,-2A,) >bd, then the plastic
neutral axis is within the web of the steel beam
follows :2
and
John Morrison Al[1+ a (A-1) (2 a 2-3 a +3) ]
M, = Y,[A,dc-A,(d,+ t,)-tw(dn+ t,)
S L[ 1+2 a (A-1) ][ 1+2 a (A-1)(4 a 2-6 a +3) ] (dn-d, - t,) ]
(b) Sections at continuous support
Introduction The moment of resistance of the section at the
Early last year Ove Arup 8- Partners were com-
missioned by the Post Office to prepare a
scheme design for a £6m. Parcel Concentra -
tion Office in Liverpool. The operational
t. "' 1
I,

T
I, al
! support has been developed using the same
procedure as that for the midspan.
The effective breadth of slab at the support
should be taken as 0.6 x the effective breadth
requirements of the building were such that a
fairly large column-free space and floor to
ceiling height were required. and in our case a
12 x 12 m grid with floor to floor heights of 7 m
for three storeys was adopted .
Composite design of the frame was con-
,,;i),. .Z,
Fig. 1
'!
'\,, J at the midspan. The effective breadth in fact
varies according to the type of loading but the
value given above is a reasonable mean. As
the concrete is always assumed to be cracked.
the only importance of this effective breadth is
for ca lculating the area of reinforcement within
sidered as on·e of the possib le solutions as it this zone.
was known that structu ral stee lwo rk wou ld be
quick to bui ld and tha t the quantity of material
Fo r most cases a va lue of 1.2 f. w ill be
required would be reduced to a minimum. su fficiently accurate for use in the overal l L b L
For design guidance. the information available
consisted of CP17 7: Part 1: 7965. which gave
frame ana lysis.
Moment of resistance
1 1
design rules for simply supported beams. This The moment of resistance of the section M,
Code would have been satisfactory if the should be found making the following assump- d,
structure had been braced by cross frames or tions. (The formulae assume the use of stan-
stair towers. In our case the overall stability d.
dard rolled sections with equal flanges.)
could have been provided by concrete stair (a) Midspan sections
towers but these would have been expensive Use methods given in CP117: Part I: 1965.
and time-consuming to construct. It was where the breadth of concrete flange should C G. of
therefore decided that a study should be made be taken as one-fifth of beam length between steel
into the possibilities of using sway frames with beam
points of contraflexure. or one -tenth of dis-
moment- resistant joints. This decision to in- tance for edge beams or the actual flange
vestigate fully continuous frame action meant width. whichever is the lesser. f!. = Y,
that we were faced with the problem of finding fy,
an acceptable design procedure. b
=

r
fy, yield stress of reinforcement

r
There was one other Code avail able for guid -
ance. CP117: Part 2: 7967. 'Beams for b, A, = area of reinforcement
bridges·. This Code gave some information for Fig. 3
continuous beam design using elastic methods.
However. it was not really suitable and it was
therefore decided to make a study of all avail - (1) If A,> PA, then the plastic neutral axis is
able technical papers to see if it wou ld be within the reinforcement and
possible to draw up our own set of design M, = A,Y,(d1 -d,)
ru les. (2) If PA,< (A,+ 2PA,) then the plastic
The design guide which fo ll ows is the resu lt of neutraf axis is within the top f lange of the stee l
this study and it has been amended from time CG of beam
- - - --'6--- - - -- -
to time as a result of our continued interest in steel beam M, = Y,[A,(d,-d,)-b,(dn-d,)
this subject. We feel that its development has
now reached a stage where we are able to give
,. (dn+ d,-2d,) ]
Conditions (1) and (2) cannot occur unless
a complete design method together with some
special stiffening is provided to stabilize the
advantages and disadvantages of this form of
web. otherwise the amount of reinforcement
construction. A Hewlett-Packard program has A, = area of steel section required would mean that the beam would fail
also been written to aid the design process. A, = area of flange the 'compact se ction· rules (see later) .
Methods of analysis Y, = yield stress of steel beam
Uw = cube strength 9 Y, (3) If A,< P(A,-2A,) then the plastic neutral
The method of frame analysis has an impor- a = -- axis is within the web of the beam. and
tant bearing on the system to be used for the Fig. 2 4 Uw
M, = Y,[A,(d, -d,)- (A,+ tw(dn-t,-d,))
analysis of the sections . For example. we (dn+d,-2d,)]
would suggest that plastic design of the frame (1) If aA,<bd, then the plasti c neutral axis dn
and its sections should only be used for con - is within slab and There are some spe cial conditions which can
tinuous beams where side sway is prevented . occur in sway frames. 3 A typical sway frame
due to the difficulty of achieving a fully rigid M r =A S YS [ d+(d,-dn) ] bending moment diagram in the lower storeys
C 2 could be as follows : (Fig . 4) .
beam to column connection and hence ade-
quate lateral stiffness. This problem will be
discussed in more detail in a later section of
this article. However. when a plastic design for
continuous beams is adopted it is important to
carry out an elastic analysis as well. as this will
be required for checking stresses at working
loads.
Where side sway can occur.' an elastic
ana lysis of the frame is preferab le; the resu lts
should then be factored by 1.75 as specified by
CPJ 77: Part 7. to give the ultimate moments
and forces. These factored values should then
be used for the determination of the section
sizes. This design procedure of applying a load
factor to the elastic analysis is the method
most commonly adopted.
For both braced and unbraced frames the
bending moment diagrams can be determined
by ignoring the inertia contribution of the Fig. 4
concrete flange in the hogging moment zones - 21
 For Zone 1 (Fig. 4 )
Momen t capacity sho ul d be calculated as for
section (a) above.
For Zone 2
Moment capacity shou ld be calculated as for
section (b) above .
For Zone 3
The moment capacity of the composite beam
adjacent to a co lumn face and subject to a
sagging bending moment should be based on
t he steel section acting together with a con-
crete flange with a width equal to that of the
column face. Condition (3) rarely occurs,
usually only in the lower storeys of multi-
storey frames.
The effective compression zone can be in-
creased as shown below.
, 'J D

Fig. 6
Typical web stiffening
<,I
· ·ffe?t ll\t
w1dtt1 Where a horizontal stiffener is provided, the capac1t1es of the section are reduced by the
upper effective limit for the force ratio becomes inter -ac tion of these forces.
0.4 . For force ra tios greater than 0.4 the plastic The ultimate shear capacity of an I beam.
axis remains approximately constant within Vu,,, can be expressed as fo llows:
the top flange of the beam, thus reducing the

Fig. 5
If necessary the concrete could be contained
to al low higher stresses or a spreader could be
incorporated to provide greater width.
Compact sections
An important restriction has to be placed on
the section sizes that can be used for con-
tinuous beams to ensure that local buckling of
the beam does not occur in hogging moment
zones. This buckling is caused by the com-
pressive stresses in the bottom flange and the
web where they are away from the stabilizing
influence of the concrete flange.
Sections which are suitable for the formation
of plastic hinges are described as compact.
Research has shown 6 that for structural steels
to Grades 43 and 50 a section can be con-
sidered compact if it satisfies the following:
For Grade 43 steel:
b The steel section must be 'compact' to ensure
Flange slenderness: - ' ,;;; 17
r,
d In both cases the stee l beam should be
Web slenderness : ,s:;70 (1-1.4 <P ) with a
Cw
lower limit of 43
For Grade 50 stee l
b,
Flange slenderness: - ,;;; 1 4
r,
d the tensi le strength of the reinforcement
Web slenderness: ,;;; 58 (1-1.4 <P) with a
lw
lowe r limit of 36
where <P. t he force ratio=;,;,
5 5
The ru les given above are more severe than
those control ling the plasticity of universa l
beams for convent iona l steelwork design. It
wil l be found that the range of universal beams
available 1s severely restricted and in many
cases it will be necessary to resort to the use of
universal column sections.
The rotational performance of slender sections
can be improved by the addition of a horizontal
web stiffener at mid-depth extending 1.5d
from the support. At the end of this stiffener
the support moment is likely to have dropped
efficiency of the extra reinforcement.
It should be noted that on ly web stiffening is Vult = j
Val - ( M-M,
Mw ) 2
required, fo r although initial ly there is a sma ll
amount of flange buckling, it is the web buck- Where V0 Tresca Criterion
ling that produces a rapid loss of strength. For
further details of this buckling refer to the M = applied moment
paper by Climenhaga and Johnson.s M, = t1 (b,-tw) (d-t,) Y,
The rules given above are satisfactory where it Mw= lwd2Y,
is intended that the full ultimate moment of 4
resistance of the support is to be developed.
However in many instances the design wrll not and M, = Mw+M,
require that the full support moment be Research by Johnson and Willminton7,B has
developed. In this case, applying the compact shown that for composite beams there is an
section rules would result in uneconomic additiona l shear capacity availab le in the
sections. hogging moment zones. This additiona l capa-
If the applied moment is less than 70 per cent city is provided by the reinforcement in the
of the ultimate moment of resistance at the negative moment regions which acts together
support. and providing that the appropriate with the steel section so that shear forces
elastic checks are carried out. then it would exceeding the plastic capacity of the web can
appear reasonable to relax the rules given be achieved. For the commonly adopted force
above. ratio values of 0.28 the increase in shear capa-
Elastic analysis check city is likely to be in the order of 10 per cent.
An elastic design check should be made for
both the midspan and support sections at
work ing loads. that sufficient rotation and plasticity can be
achieved before the maximum moment is
checked to ensure that the elastic stress does reached. If this ro tation cannot occur, then the
not exceed 0.9 yield stress . At the midspan the section is likely to farl by buckling.
concrete stress should be checked to ensure Usual ly this additional shear capacity, due to
that it does not exceed one-third of the speci- the dowel action of the reinforcement. w ill not
fied cube stress, although this condition 1s be requireci and a simple check us ing the
Tresca Criterion will be sufficie nt.
unlike ly to control the design . At th e support
shou ld be checked to ens ure that this does not
Con,pu~11t1
exceed 0.6 yie ld stress of the reinfo rcement. hcdrn r. u, ve
In virtua lly al l cont inuo us composite beam
design the 0.6 yield stress in the reinforcement
or the 0.9 yield stress in th e stee l beam w ill be
the contro lli ng factors in the design.
To enable the elastic properties of the section
to be calculated. a knowledge of the modular
ratio m wi ll be required . Part 1 of CP717
recommends m= 1 5 while Part 2 recommends
that m=83vuw.
For a midspan section it can be shown that
there is little change in stress in the critical
bottom flange for a range of modular ratio
between 8 and 16. The stress change is less
than 5 per cent and it is therefore recom -
mended that m= 1 5 should be adopted. 0 I ()

to 70 per cent. Horizontal stiffeners should
always be provided for the column webs 1n
line with the bottom flange of the beam to
prevent local buckling of the column web.
Without additional stiffening the rules given
above effectively limit the value of force ratio
to 0.28, because at the limit of d/tw 36 there
are on ly a few ro lled sections tha t can comply
22 w ith this cri teria.
For the support condition the modular ratio is
not included in the calculations as the con-
crete is assumed to be fully cracked.
Shear
For simply supported beams the shear force
should be supported entirely by the web of the
steel beam.
For cont inuous beams th e situation is sligh tl y
mo re co mp lex in that the shea r and momen t
Fig. 7
Bending -s hear interaction curve
Design of shear connectors
Number of connectors
For simply supported beams the design load
per stud Pc should be the va lue given in
CP 11 7 wh ich is 80 per cent of its ulti ma te
capacity. It has been shown 14 that if this value
of Pc is adopted then the amount of slip will
be limited to a value which will not impair the
ultimate moment of resistance.
For uniformly distributed loads the number of
connectors required can be determined as
follows:
When plastic NA is within depth of steel
. h b . Fee
section t e num er required= Pc' where Fee
is compressive force in slab at ultimate load.
Or when plastic NA is within depth of slab the
_ Fee Fst
number required= Pc or Pc" where Fst is
tensile force in steel beam at ultimate load.
The number of connectors given above repre-
sent the tota l required between the points of
zero and maximum shear. Thus for a simply
supported beam with a uniformly distributed
load. twice the value given above will be
required over the full length of the beam.
The connectors may be spaced evenly along
the length of the beam.
Where point loads predominate. it is necessary
to divide the shear force diagram into sections
between the points of application of the loads
and to calculate the total shear force for each
zone. The number of connectors required can
then be calculated and these should be evenly
distributed along the section considered.
For continuous beams the connectors in the
hogging moment zones are likely to be in an
area of cracked concrete . For this reason it has
been found that an acceptable design load for
the studs to control slip is 80 per cent of the
value given in CP7 77. i.e. 64 per cent of the
ultimate capacity.9
The number of connectors required should be
found by calculating the horizontal shear
force. which is equal to the tensile force
generated in the slab reinforcement over the
support.
. A,~,
Thus the num b er requ ired N= 0 _8 Pc
Spacing
A lthough the method of determining the hori-
zonta l shear force is well known, there appears
to be far less agreement as to the manner in
which shear connectors should be distributed
along a composite beam.
Johnson. Greenwood and Van Dalen 9 recom-
mend that the number of connectors in the
hogging moment zone. as found above. should
be equally spaced between the points of
maximum hogging moment and contraflexure.
N,
horizonta l shear forces between the points of
maximum and minimum moment and that
these connectors should be equally spaced
along the beam.
N = (A,f., + Feel
P,
Fig. 9
There is little difference in the final result
between the two methods. The latter system is
slightly more convenient in practice and is
therefore recommended.
In all cases it is recommended that the spacing
of the connectors should not exceed three
times the slab thickness or 600 mm. whichever
is the lesser.
If a single line of shear connectors is al l that is
required. these should be staggered on either
side of the centre line to reduce the possibility
of longitudinal splitting of the slab.
Height of shear connectors
In sagging moment zones the height of the
studs should extend 25 mm into the com-
pression zone and never be less than 50 mm
high to ensure adequate resistance against
uplift.
For hogging zones uplift is no longer a prob -
lem. The problem is now one of loss of
efficiency, due to slip. as the connectors are
embedded in a zone of cracked concrete. In
this zone it is recommended that the shear
connectors shou ld project at least 25 mm
above the lowest leve l of re inforcement.
However for practical reasons the height of
connectors should be the same throughout the
length of the beam and therefore the height of
connectors determined for the sagging zones
will usually control. (See Table 1 ).
Design of transverse reinforcement
in slab
Using the method set out in CP7 77: Part 7. the
following checks should be made :
The shear force Q in N/mm2
Ne x load in N on one shear connector
0 = at ultimate load
Longitudinal spacing of c_o_n_n_e_c_t_
o-rs~i_n_m
_m
This force should not exceed the lower value
of either:
(1) the shear resistance per mm run of
beam=0.23 Lsvuw+A/,,n
or
(2) 0.62 Lsvuw
Where Ls= shear perimeter of the connectors.
but:} 2 x slab thickness ofT beams
::} slab thickness for L beams
n= number of times the lower level of
reinforcement intersects the shear
perimeter Ls.
For T beams n= 2. for L beams
n=l.
The amount of transverse steel A, in the bottom
Q
of the slab should not be less than 41Y, mm 2 per
mm run of beam.
Reinforcement provided to comply with this
requirement can also be assumed to resist
transverse slab moments provided that the
minimum quantity in the bottom of the slab is
maintained.
This rule is especially important to prevent
splitting of the slab when high shear forces
occur.
An alternative method which results in a reduc-
tion in the quantity of transverse reinforcement
has been proposed by Johnson. 11
Deflection
Although composite design of a beam has
many advantages. such as a reduction in the
depth of steel beams. the use of these
shallower beam depths means that deflection
has to be carefully considered.
For continuous beams the deflection is un-
likely to exceed 20 per cent of the deflection
for a simply supported beam with the same
moment of inertia. Even if the support moment
were reduced to 50 per cent of the midspan
value the deflection would still only be 30 per
cent of the simply supported condition. A
further benefit of continu ity is that shrinkage
and creep of the concrete f lange are unlikely to
give rise to a serious deflection problem and
can therefore be neglected.
The most serious deflection occurs with
simply supported beams and for these special
care should be taken . Not only are dead and
live load deflections important. but also the
creep and shrinkage deflection. which in some
circumstances can be as much as 100 per cent
of the dead load deflection' 2.

Table 1
Design values of shear connectors for different concrete strengths

Design values of connectors for

Type of connector Connector material concrete strengths U w N / mm 2

Headed studs 20 30 40

Diameter mm Height mm Load per stud kN

A,f,, Fee 25 102 Minimum yield 117 135 153

N , = -- N,=- 22 102 stress 386 N/mm 2 97 112 126
0.8 Pc Pc 19 102 77 88 100
19 76 Minimum tensile 66 76 87
16 76 stress 494 N/mm 2 56 65 74
Fig. 8 13 64 35 42 47

Bars with hoops BS4360

For the sagging moment zone the number of
Grade 43
Fee . 50x 38 X 200 380 580 780
connectors is Pc which should be equally

distributed between midspan and the point of Channels

contraflexure.
127 x 64 x 4.5 Kg x 152 BS4360 22 3 260 297
On the other hand . Yam and Ch apman 1 0 102x 52 x 3.18 Kg x 152 Grade 43 206 242 280
recommend that the number of connectors 76 x 38x 2.04 Kg x 152 194 230 266
sh ould be determined using the sum of the 23
 For the calculation of the deflection it is
recommended that the effective modulus
method should be used. This is a simple and
well-proven method and is used in both the
CEB and CPI 10 recommendations. It has been
shown by Knowles 15 that the results do not
vary significantly from the more 'exact'
methods.
The modular ratio adopted should be modified
to all ow for the effects of shrinkage and creep,
either by using the CEB graphs or by adoption
of the simpler values given in Appendix A of
the explanatory handbook to CPI 10. The final
calculation can then be carried out using the
curvature method together with the K factors
given in Table 1 of the explanatory handbook.
The construction sequence also has an impor-
tant bearing on the final deflections. If the
beams are not propped during the casting of
the slab, deflection due to this self-weight of
wet concrete actir,g on the steel beam alone
should be allowed for, possibly by pre -
cambering the beams. If the beams are
propped, usually at one-third points, the con-
crete is al lowed to harden, before the props are
removed after 14 days. The total load is now
resisted by the composite beam and the deflec-
tion is therefore reduced. A further advantage
of this second method is that it reduces the
initial stressing of beams due to self-weight of
the concrete and in certain instances bene -
ficial effects can be achieved by an initial
Jacking up of the beams.
However, jacking up of the beams on site is
difficult to control in practice. It may also be
found that in many instances even propping 1s
an unnecessary inconvenience. The effects of
all of these factors will need to be taken into
account for determination of the beam size.
Beam to colum connections
The design of the beam to co lumn connections
could be looked upon in many respe cts as the
starting point of the design. For until a
decision has been made whether to adopt
Terismn bol1~ H.S F.G
Neu 1ral ax is
Fig . 11
Rigid connection
Fig.12
Partial fixity Joint
24
simply supported, full fixity or some inter-
mediate partial fixity, it is impossible to
analyze the frame or determine the beam
moments.
In an earli er section of this report the effect of
ana lysis and section properties was discussed.
It is now intended to take this a stage further
and to discuss the impli ca ti on of various
assumptions about the beam to column con-
nections.
Simple support connections
These connections should on ly be used for
single bay structures. Simple supports can be
provided in multi-bay structures, provided that
a joint is formed in the concrete slab at all lines
of support, but this is not recommended.
If reinforcement is provided in the slab over the
support then the section is no longer simply
supported and in fact some degree of partial
fixity has been introduced. Thus depending on
the loading, span and depth of beam, the mid-
span is likely to be oversized while the support
may be overstressed.
Fig.10
Simply supported connection
using bla ck bolts
- ~ Shea, bol 1s H S. F G.
~ fully !Orq ued
H.S. F.G bol1 s
Rigid connections
If a rigid connection is provided either by
welding or the use of high strength friction
grip bolts, and provided that the section is
compact then the greatest economy can be
achieved by taking advantage of fu lly con-
tinuous design.
However, welded joints are likely to be expen-
sive. Bolted Joints also have their problems due
to practical difficulties of prov1d1ng sufficient
bolts to effectively develop the full plastic
moment. There is also the problem of service-
ability at work ing loads, for in an ideal condi-
tion midspan and support moments should be
of the same order. However, at working load
the elastic stress in the steel must be limited to
0.9 Y, and the stress in the reinforcement to
0.6 Y,. These two conditions are always likely
to be the controlling factors for determining
the moment of resistance at the support.
Assuming that the elastic stresses at the
support are within acceptable limits, 1t may still
be difficult to provide a bolted connection with
sufficient streng th to take the balance of the
forces after making al lowan ce for the reinforce -
ment. (See Fig.11) .
Mu,, (Bolts)= Mu1,-M,e,n1
Partial fixity
It has been shown above that the simply
supported design can lead to uneconomic
section sizes, while rigid design gives support
moments which for all but a very few instances
are difficult to achieve. Yet as we have seen
earlier there are several distinct advantages to
be gained from continuity, i.e.:
(a) The section sizes can be reduced.
(b) The deflections can be minimized.
(c) The reduction in deflection makes the use
of HYS more advantageous.
The desirable joint is therefore one which lead s
to a fairly simple and conventiona l bolting
arrangemen t while providing a reasonable
support moment acting in con Jun ction with
the reinforcement.
,,"'
1 :•
+·
I
• tI ,...•.... -4•t__
tl ~:+ "
--t-
• ...,. t
+
• •. ,
Nl•u11.1I
1,1-.
ii i>'
++
+ --,- I
,1
For the purpose of providing a partial fixity
Joint. either of the details shown in Figs. 11
and 1 2 can be used.
The details shown in Fig. 1 2 have been
developed by Johnson and Hope-Gill, 1 3
although somewhat clumsily. It is simple to
fabricate and the small gap between the beam
and column ensures adequate rotational
capacity.
Partia l fixity Joints should not be used in sway
frame s as these Joints can only be used where
stability is provided by some alternative
system.
The stresses set up by the se lected support
moment must be checked to ensure that at
working loads the stresses in the beam and
reinforcement are not exceeded. In the case of
the detail shown in Fig . 12, it is also important
to check that slip of the bottom flange bolts
does not occur at working loads.
Design procedure
(1) Derive ultimate and elastic bending
moment and shear force diagrams.
(2) Determine slab depth. using span/dep th
ratio for con tinu ous slab.
(3) Ca lcu late effective flange w idth at both
midspan and sup port.
(4) Select a suitab le stee l beam. approxi-
M method
mately Z=TI r-;;;
(5) For the sagging moments calculate:
(a) ultimate moment of resistance
(b) the elastic properties of the section.
(6) For the hogg ing moments. repeat (a)
and (b).
(7) Check th at the section is compact.
(8) Check elastic stresses in steel beam and
conc rete. midspan section . at work in g
loads .
(9) Check elastic stresses in stee l beam and
reinforcement. sup port secti on at work-
ing loads.
(10) If stre sses at support are a major prob -
lem. cons ider the possibility of simply
supported des ign or some form of partial
fixity.
(11) Check shear at supports for ultimate load
condition.
(12) Design moment -resistant joint between
beam and column.
(13) Ca lculate longitudinal shear force at
ultimate load. for calc ul ation of shear
connectors.
(14) Select type of shear connector. calcula te
number and spacing .
(15) Check long itudinal shear force in co n -
crete.
(16) Design transverse reinforceme nt in slab.
(17) Ca lcu late defl ecti on taking account of
creep. shrinkage. constr ucti on seq uence
and method of propp ing.
(18) Check stresses in steel beam under co n-
struction loads.
Hewlett-Packard programs
One of the most daunting aspects of composite
beam design is the labo riou s nature of the
calcu lations and for continuous composite
beams the amount of work is more than
doub led. To carry out a comp lete analysis of a
compos ite beam. including elas ti c and deflec-
tion checks. could easily take one day. There -
fore this design gu id e ha s been embodied into
a comp uter program. enabling the eng in eer to
exp lore all possible variations before making a
final decision on the section sizes.
A program in three main parts. based on the
use of structura l rolled steel sections. has been
prepared.
Pare 1
(a) Determines the ultimate moment of resis-
tance at the midspan and support
(b) Checks that the section is compact 1f the
beam is cont inuous
(c) Checks th e elastic properties of the sec-
tion at both midspan and supports giving :
neutral axis depth
(equivalent steel) moment of inertia
Z concrete top
Z stee l top
Z co ncrete bottom
(d) Checks at midspan that the concrete stress
< 0.33 Uw and that the stee l stre ss
< 0.9 Y,
(e) Checks at support that the stress in the
reinforcement < 0.61,, and that the stress
in the beam < 0.9 Y,
(f) Checks the shear capacity of the web.
Pare 2
(a) If required. checks the add iti ona l shear
capaci ty by the Johnson, Willminton
method 8 .
(b) Gives the number and spac in g of shear
co nn ec tors
(c) Gives the area of transverse re inforce ment
in the slab
(d) Given bolt size. designs the joint for either
simply supported. partially fixed or fully
fixed cond itions and gives the thickness
of th e end plate.
Pare 3
(a) Checks deflection using CP110 method
(b) Checks additional deflection due to creep
and shrinkage using the effe cti ve modulus
(c) Checks deflection due to var i o u s
sequences of construct ion. i.e. free span.
single midspan prop. or props at one-
thi rd points.
Conclusions
It is fa irly wel l-establ ished that the composite
design can result in a 30 per cent saving in
structural stee lwork as compa red with a non-
composite design. When contin uous compo-
site beams are adopted there can be a furthe r
10 per cent saving . Togeth er w ith th e reduction
of beam depth and deflecti on. and by vi rtue of
thi s reduction in deflection. a more advan -
tageous use of high yield stee l can be made.
Full fixity of the beam co lumn is possible for
beams with a depth of up to 550mm . For
beams deeper than 550 mm. some form of
partial fixity is desirable. .
Without special stiffening of the web and beam
flanges. it is important to limit the force ratio to
0.28. otherwise the ·compact section· rules
wil l be failed.
For preliminary des ign s the steel section size
ca n be found by determin ing the elastic
modulus using the midspan bending moment
and the allowable bending stress which should
be factored by 1 .4.
It has not been poss ibl e to deal with th e many
other possible variat ions o f compos ite beam
const ru cti on. but as food for th ough t the
following comments may be interesting.
Haunched beams
Th e formation of a co ncrete haun ch in the slab
greatly improves the efficiency of simply
supported beams due to the increase in lever
arm . However. fo r normal building frames. the
Fig. 13
Haunched beam
L., llle,1,"'''""""" ,,,,,,
Fig.15
Formed meta l deck
advantages are like ly to be eliminated by the
expense of the shuttering.
Included in this category are the steel sections
with un sym metri ca l flanges. although these
aga in are likely to be expensive to fabricate
and are only suitab le for simply supported
co nditions.
Precast concrete decks
These are again ideal for simply supported
beams. but care must be taken to ensure that
there is sufficient bearing on the flange for the
slabs and at the same tim e suffi cien t width of
concrete to enclose the shear connecto rs.
Ca re should also be give n to providing suffi-
cient transverse reinforcement to prevent
splitting of th e slab.
Precasting could be an effective way of
achieving a haun ched beam.
In
~1t11
Fig.14
Precast slab
Formed metal decks
This type of decking is id ea l fo r both sim ply
supported and continuous designs . The shear
studs should be welded to the beam through
the decking to ensure the beam stiffness and
to minimize slip . For rib heights of less than
40 mm there is no sign ifi cant loss of strength .
However. with rib heights greater than 40 mm
there 1s a possibil ity of shear occurring across
the ribs. 17 (See Fig.15).
Prestressing
Thi s can be ach ieved in many ways; by con-
ventional stressing methods. the use of Pref/ex
beams: or sys tems which in vo lve Jacking. Th e
principle advantage is an in crease of the elastic
range of th e steel beam. Preflexing and pre-
str essing are usual ly con fined to use on simp ly
supported beams. although the Jacking system
can be useful with con tinuou s beams.
Castellated beams and lattice trusses
Considerable increases in rigidity can be
achieved. Due to the problems of web buck-
ling these are not suitable for continuous beam
design. The most efficient design will result
where the shear forces are low.
25
 References
(1) STEVENS. L. K. Elastic stability of prac-
tical multi-storey frames. Proceedings of 1he
lns111ution of Civil Engineers. 36 (Jan.). pp.
99-117.1967.
(2) JOHNSON. R. P. et al. A plastic com -
posite design. Proceedings of 1he lnsuwuon of
Civil Engineers. 32 (Oct.). pp. 198-209. 1965.
(3) DANIELS. J. H. et al. Behaviour of com-
posite beams to column Joints. ASCE Pro -
ceedings. Journal of the Strucwral Division.
96 (ST3). pp. 671-685. 1970.
(4) BRITISH STANDARDS INSTITUTION.
CP117: Part 1 : 1965. Composite construction
in structural steel and concrete. Part 1 : Simply -
supported beams in buildings. BSI. 1965.
(5) BRITISH STANDARDS INSTITUTION.
C P117 : Part 2 : 1967. Composite construction
in structural steel and concrete. Part 2: Beams
for bridges. BSI. 1967.
(6) CLIMENHAGA. J. J . and JOHNSON.
R. P. Local buckling in continuous composite
26
beams. Strucwral Engineer. 50 (9). pp. 367-
374. 1972.
(7) JOHNSON. R. P. and WILLMINGTON.
R. T. Vertical shear in continuous composite
beams. Proceedings of the lnstiwtion of Civil
Engineers. 53 (Sept.). pp. 189-205. 1972.
(8) JOHNSON. R. P. and WILLMINGTON.
R. T. Vertical shear strength of compact com-
posite beams . Paper 7418S. Proceedmgs of
the lnsuwuon of Civil Engmeers, Supplement.
1972.
(9) JOHNSON. R. P. et al. Stud shear-
connectors in hogging moment regions of
composite beams. Strucwral Engineer. 47 (9).
pp. 345-350. 1969.
(10) YAM. L. C. P. and CHAPMAN. J. C.
The inelastic behaviour of continuous com-
posite beams of steel and concrete. Proceed-
ings of the lnsuwuon of C1vt! Engmeers. 53
(Dec. ). pp. 487- 501. 1972 .
(11) JOHNSON . R. P. Longitudinal shear
strength of composite beams. AC/ Proceed-
ings. 67 (6). pp. 464-466. 1970.
(12) AMERICAN CONCRETE INSTITUTE.
Publication SP-27. Designing for effects of
creep shrinkage temperature in concrete
structures. Effects of differential shrinkage and
creep on a composite steel-concrete struc-
ture. by F. Roll. pp. 187-214. ACI. 1971.
(13) JOHNSON. R. P. and HOPE-GILL. M .
Semi-rigid Joints in composite frames. Inter-
national Association for Bridge and Structural
Engineering. Ninth Congress. IABSE. 1973.
(14) CHAPMAN. J. C. Composite construc-
tion in steel and concrete - the behaviour of
composite beams. Strucwral Engmeer. 42 (4).
pp.115-125.1964.
(15) KNOWLES. P. R. Composite steel and
concrete construction. Buttervvorths. 1973.
(16) SHERBOURNE. A. N. Bolted beam to
column connexions. Strucwral Engmeer. 39
(6). pp. 203-210. 1961.
(17) FISHER . J . W. Design of composite
beams with formed metal deck. Engmeering
Journal of the Amencan lnst11Ute of Steel
Construction. 7 (3) . pp. 88-96. 1970.